JP2012088758A - Storage device and control method of storage device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a storage device and a control method of the storage device such that even if writing of data to a nonvolatile memory is interrupted, the nonvolatile memory can be started up next time without causing format illegality.SOLUTION: The storage device includes a nonvolatile memory management service 1 (control mechanism), an external storage device 2, and a nonvolatile memory 3. The external storage device 2 has a nonvolatile memory use size 21 storing a nonvolatile memory use size, and a Checksum value 22 storing a check sum. The nonvolatile memory 3 has a plurality of Entry's 31. The plurality of Entry's 31 each have respective items of a Signature 311 indicative of whether each Entry is effective, a Status 312, a data length 313, and a data part 314. When the writing is restarted after interruption, the check sum is verified again to determine whether the interruption is caused by a fault of the nonvolatile memory 3.

Description

  The present invention relates to a storage device and a storage device control method, and more particularly, to a storage device and a storage device control method capable of checking the reliability of a storage medium even when a power failure occurs.

In recent years, server platforms such as information processing devices have defined standard interfaces for the purpose of storing BIOS (Basic Input / Output System) and OS (operating system) configuration information, and the BIOS is non-volatile. Provides services to add and delete data from memory.
The nonvolatile memory stores information and failure logs necessary for starting up the BIOS and OS, and other important information necessary for system operation.
In addition, since the non-volatile memory may break down due to deterioration due to long-term use or the like, conventionally, a technique such as duplicating the non-volatile memory has been proposed.
However, when power is cut off during writing, there is still a problem that the data cannot be matched and the checksum check becomes inconsistent. The present invention also has a problem of solving this problem. Yes.

  As a known technique prior to the present invention, for example, in Patent Document 1, when data is recorded in a non-volatile storage device, how far the data can be trusted even when the data writing process is interrupted. A data writing method and apparatus capable of easily discriminating the above are disclosed. Specifically, before the start of data writing to the non-volatile storage means and during the writing process, the presence or absence of a factor that interrupts the data writing is detected, and the detection result is recorded in an area for recording a marker bit provided in advance. Appropriate data shall be added and written. It is assumed that the range of reliable data can be determined by detecting the change point of the marker bit after restarting.

  Further, for example, Patent Document 2 discloses a non-volatile memory management method capable of reading normal stored data and recovering the error even if a write error occurs due to a power failure at the time of data update, and can be used continuously. Has been. Specifically, the user data is written to the user data area, the user data array indicated by the user data array reference information is read when the writing is successful, the validity check information calculation writing process is executed, and the written data and the read data are collated. In the case of coincidence, it is determined whether or not writing to the user data array is completed by the number of user data array reference information. If the data validity does not match, the data is changed to a new user data array, error recovery is performed, the data is re-executed after being written to the user data area, and is continuously used.

  Further, for example, Patent Document 3 discloses a header conversion device that reduces the time required for recalculation of checksum performed at the time of header conversion. Specifically, when a packet having at least one header provided with a plurality of fields including a checksum field is input, conversion means for converting a value of a predetermined field excluding the checksum field of the header; A difference value storage means for storing a difference value of a checksum calculated using values before and after conversion of the predetermined field converted by the conversion means; a packet of which the value of the predetermined field is converted by the conversion means; Rewriting means for rewriting the value of the checksum field of the header to a new value using the difference value stored in the difference value storage means is provided.

  Further, for example, in Patent Document 4, the reliability of a nonvolatile memory in which writing is frequently performed can be maintained, the destruction of the nonvolatile memory can be prevented and the life can be extended, and the life time can be estimated from the appearance. A technique that can be applied is disclosed. Specifically, a data storage nonvolatile memory for storing data, a checksum storage nonvolatile memory for storing a checksum of the data storage nonvolatile memory, and data in each nonvolatile memory A central processing unit having a function of controlling reading / writing of data and a function of calculating a checksum from a non-volatile memory for storing data, and a display unit for displaying a checksum error, the number of times of writing, and a memory life, A predetermined digit below the checksum calculated by the central processing unit is used as a checksum storage address of the checksum storage nonvolatile memory, and a write mark or a write count is stored in the checksum storage address.

JP 2000-132464 A JP 2001-331382 A Japanese Patent Laid-Open No. 2003-124977 Japanese Patent Laid-Open No. 7-141899

By the way, as described above, the nonvolatile memory stores information necessary for starting up the BIOS and OS, and other information (for example, a failure log), that is, important in operating the system. Information is stored.
However, in the conventional storage device control method described in the background art above, when writing to the nonvolatile memory is interrupted due to a sudden power interruption or a FATAL failure, the nonvolatile memory is expected. There is a problem that the data may be stored in an external format, and there is a case where the system cannot be started up unless the nonvolatile memory is completely erased.

  For example, since data to be written to the nonvolatile memory has a variable length, the data length must be stored in the data. However, if writing is interrupted while the data length is not stored due to a sudden power failure or a FATAL failure, the format will not be preserved at the next boot, so the BIOS will save the contents of the nonvolatile memory. Will be determined to be fraudulent and stop the system startup.

  In addition, the BIOS has a checksum for diagnosing non-volatile memory, but if the power is cut off without performing the checksum update process, the checksum will not match at the next startup. However, there is a problem that the nonvolatile memory is determined to be illegal. The reason why such a situation occurs is that the checksum is updated immediately after the data update, but the checksum is not updated when the power is suddenly turned on. As a countermeasure to solve this problem, there is a method of dividing the nonvolatile memory and taking the checksum, but the nonvolatile memory is considered invalid for the block that has not been updated with the checksum due to the power failure. It is impossible to improve until it is done.

  In addition, in the known techniques described in Patent Documents 1 to 4, general data managed by the user is a diagnosis target by checksum and is a recovery target, but is stored in a nonvolatile memory. It was insufficient as a fundamental diagnosis and recovery process for information necessary to start up the existing BIOS and OS from interruption and other information (for example, failure log).

  On the other hand, in the storage device according to the present invention, the first storage medium for storing data and verifying the data using the checksum value of the stored data, and the block size and the checksum value of the stored data are stored. A second storage medium that controls the first and third storage media, and writing to the first storage medium is interrupted due to a sudden power failure or a FATAL failure. Even in this case, it is important to refer to the information stored in the second storage medium so that the apparatus can be normally reused without causing a format error at the next startup of the apparatus. It is said.

The present invention has been made in view of the above-described conventional problems, and even if data writing to the nonvolatile memory is interrupted, the next startup will start without causing an illegal format of the nonvolatile memory. It is an object of the present invention to provide a storage device and a control method for the storage device.
Another object of the present invention is to enable checksum verification by treating only valid parts as valid values when checking the normality of data stored in a nonvolatile memory, even if a power failure occurs. Another object of the present invention is to provide a storage device and a storage device control method capable of performing reliability check of a nonvolatile memory.

  In order to solve the above-described problem, a storage device according to the present invention stores first data, attribute information of the data, and attribute information of the entry in an entry constituting one storage unit. A second storage medium that stores a use amount of the first storage medium up to the present time, and a checksum value for storage data in the first storage medium indicated by the use amount And control means for controlling the storage contents of the first and second storage media based on a given command.

  The storage device control method according to the present invention stores given data, attribute information of the data, and attribute information of the entry in an entry constituting one storage unit of the first storage medium. Storing the amount of use of the first storage medium up to the present time and the checksum value for the storage data in the first storage medium indicated by the amount of use in the second storage medium And a control step for controlling the storage contents of the first and second storage media based on a given command.

  As described above, according to the storage device of the present invention, the hardware includes the first and second storage media, and the second storage medium includes the first storage medium (for example, non-volatile storage). Memory) up to the present time and the checksum value for the storage data in the first storage medium indicated by the usage amount are stored, so that sudden power interruption or FATAL failure occurs. Thus, even if the writing to the first storage medium is interrupted, the format error will not occur at the next startup of the apparatus by referring to the information stored in the second storage medium. There is an effect that the apparatus can be normally reused.

It is a block diagram which shows the whole structure of the memory | storage device which concerns on embodiment of this invention. It is a flowchart figure which shows the process sequence at the time of the variable new creation process of the memory | storage device which concerns on embodiment of this invention. It is a flowchart figure which shows the process sequence at the time of the variable deletion process of the memory | storage device which concerns on embodiment of this invention. It is a flowchart figure which shows the process sequence at the time of the variable update process of the memory | storage device which concerns on embodiment of this invention. It is a flowchart figure which shows the process sequence at the time of the access request process of the memory | storage device which concerns on embodiment of this invention. It is a flowchart figure which shows the process sequence in the process at the time of starting of the memory | storage device which concerns on embodiment of this invention.

The present invention is intended to preserve the data stored in the nonvolatile memory. Specifically, even when the value is stored in the nonvolatile memory, even if the power is turned off, a format error is caused in the next boot. So that the system can be up and running.
Also, in checking the normality of data, it is possible to check the reliability of the non-volatile memory even if the power is cut off by enabling only valid parts to be handled as valid values.
Therefore, in the present invention, the storage format of the nonvolatile memory is defined, and the used size and the checksum value of the nonvolatile memory are stored in an external storage device installed separately from the nonvolatile memory. Then, a storage format of the nonvolatile memory and a predetermined access procedure to the nonvolatile memory are executed.

Embodiments of a storage device and a storage device control method according to the present invention will be described below in detail with reference to the drawings.
FIG. 1 is a configuration diagram showing the overall configuration of a storage device according to an embodiment of the present invention.
The storage device according to the present embodiment shown in the figure includes a nonvolatile memory management service 1 (control means), an external storage device 2 (second storage medium), and a nonvolatile memory 3 (first storage medium). .
The external storage device 2 includes a nonvolatile memory usage size 21 that stores a nonvolatile memory usage size, and a Checksum value 22 that stores a checksum.
The nonvolatile memory 3 includes a plurality of entries 31 (entries).
In the present invention, generally, any type of readable / writable storage medium can be used as the nonvolatile memory 3.

  Each of the plurality of entries 31 includes a signature 311 (signature storage item) indicating whether the entry is valid, a status 312 (status storage item), a data length 313 (item storing data length), and a data portion 314. (Items for storing data).

In this embodiment, the nonvolatile memory management service 1 controls the external storage device 2 and the nonvolatile memory 3 in response to a request from a higher-level management system (not shown). A control system (not shown) may be installed below the volatile memory management service 1 so that the control system controls the external storage device 2 and the non-volatile memory 3 in response to a request from the non-volatile memory management service 1. Is possible.
Further, although the external storage device 2 is here one component of the storage device according to the present invention, it may be an external storage device installed at an arbitrary point including a remote place.

Hereinafter, the operation of the storage device of this embodiment will be described.
FIG. 2 is a flowchart showing a processing procedure at the time of new variable creation processing of the storage device according to the embodiment of the present invention.
Hereinafter, with reference to FIG. 1, a processing procedure at the time of new variable creation processing of the storage device according to the present embodiment will be described using the flowchart shown in FIG. 2.
Here, a case where the nonvolatile memory management service 1 receives a new data creation instruction from an external host system will be described.
First, in step A 1, the nonvolatile memory management service 1 creates a new entry with Signature as invalid at the end of the nonvolatile memory 3.

Next, in step A2, the nonvolatile memory management service 1 stores the newly created data and its data length in the created entry (more specifically, the data length of the newly created data is entered in the data length 313 as the entry item). The data is stored in the data portion 314 of the data).
Next, in step A3, the nonvolatile memory management service 1 sets the status of the created entry to “ADDED” in the character string (more specifically, stores “ADDED” in the Status item 312 of the entry item).
Next, in step A4, the nonvolatile memory management service 1 validates the signature of the created entry (more specifically, for example, sets a flag indicating validity in the signature 311 of the entry item).

  Finally, in step A5, the non-volatile memory management service 1 stores the non-volatile memory usage size and the checksum value in the external storage device 2 (more specifically, the non-volatile memory usage size 21 is non-volatile. The memory use size is stored in the checksum value 22 and the checksum value calculated in the memory use size range of the storage area of the nonvolatile memory 3 is stored).

FIG. 3 is a flowchart showing a processing procedure during variable deletion processing of the storage device according to the embodiment of the present invention.
Hereinafter, with reference to FIG. 1, a processing procedure at the time of variable deletion processing of the storage device according to the present embodiment will be described using the flowchart shown in FIG.
When the nonvolatile memory management service 1 receives a data deletion instruction from an external host system, the following processing is started.

First, in step B1, the non-volatile memory management service 1 in the non-volatile memory 3 “DELETED” of a character string indicating invalidity (and free space) in the status item 312 of the entry holding the data to be deleted. Is stored.
Next, in step B2, the non-volatile memory management service 1 stores the non-volatile memory usage size of the data to be deleted in the non-volatile memory usage size 21 of the external storage device 2, and stores the checksum value described above in the external The checksum value 22 of the storage device 2 is stored.

FIG. 4 is a flowchart showing a processing procedure during variable update processing of the storage device according to the embodiment of the present invention.
Hereinafter, with reference to FIG. 1, a processing procedure at the time of variable update processing of the storage device according to the present embodiment will be described using the flowchart shown in FIG. 4.
When the nonvolatile memory management service 1 receives a data update instruction from an external host system, the following processing is started.
First, in step C1, the nonvolatile memory management service 1 deletes data according to the procedure shown in FIG. 3 for the entry that holds the data to be updated.
Next, in Step C2, the nonvolatile memory management service 1 writes the update data to the entry.

FIG. 5 is a flowchart showing a processing procedure during access request processing of the storage device according to the embodiment of the present invention.
Hereinafter, a processing procedure at the time of access request processing of the storage device according to the present embodiment will be described using the flowchart shown in FIG. 5 with reference to FIG.
Here, a process when the nonvolatile memory management service 1 receives a data read request will be described.
First, in step D1, the nonvolatile memory management service 1 sets the search position in the top area (not shown) of the nonvolatile memory 3.

Next, in step D2, the nonvolatile memory management service 1 searches for entries in order from the search position. First, it is verified whether or not the signature 311 of the current entry is invalid. If the signature 311 of the current entry is invalid, it is determined that there is no target data variable, and the process ends as a case of Not Found. If Signature 311 of the current entry is not invalid, the process proceeds to step D3.
In step D3, the nonvolatile memory management service 1 verifies whether the signature 311 of the current entry is the target data. If the signature 311 of the current entry is the target data, the nonvolatile memory management service 1 sets the target entry to the current entry. It is determined that the data is stored, and the process ends as the case of Found. If the signature 311 of the current entry is not the target data, the process returns to step D2 to repeat the search from the next entry.

FIG. 6 is a flowchart showing a processing procedure in processing at the time of starting the storage device according to the embodiment of the present invention.
Hereinafter, with reference to FIG. 1, a processing procedure in processing at the time of starting the storage device according to the present embodiment will be described using the flowchart shown in FIG. 6.
This processing procedure at the time of start-up is executed when processing that has been interrupted due to a system failure or the like is resumed.
First, in step E1, the nonvolatile memory management service 1 acquires the contents of the nonvolatile memory usage size 21 (nonvolatile memory usage size) and the contents of the checksum value 22 (checksum) from the external storage device 2. To do.

Next, in step E2, the non-volatile memory management service 1 calculates a checksum for data stored in the storage area of the non-volatile memory 3 within the acquired non-volatile memory usage size range.
Next, in step E3, the nonvolatile memory management service 1 verifies whether or not the checksum calculated in step E2 matches the contents (checksum) of the checksum value 22 acquired in step E1. If the checksums do not match, the process proceeds to step E4. If the checksums match, the process proceeds to step E6.
In step E4, the non-volatile memory management service 1 suspects that the power supply was cut off during the previous data deletion (that is, the status item 312 of an entry is changed from the character string “ADDED” to the character string The numeric data (partial checksum) indicated by the character string “ADDED” and the numerical data indicated by the character string “DELETED” The difference value from the (partial checksum) is added to the calculated checksum.

In step E5, the non-volatile memory management service 1 verifies again whether the contents (checksum) of the checksum value 22 acquired in step E1 can be matched if the added checksum is used. (The reason for performing such verification will be described later). If the checksums cannot be matched, the process proceeds to step E7. If the checksums can be matched, the process proceeds to step E6.
In step E6, the non-volatile memory management service 1 determines that the non-volatile memory 3 is normal and initializes an unused area. Thereafter, the process ends (normal end).
In step E7, the nonvolatile memory management service 1 determines that a failure has occurred in the nonvolatile memory 3, and ends the process (abnormal termination).

Note that the procedure shown in the flowchart of FIG. 6 is based on the premise that the nonvolatile memories are used in order from the front. In this way, the nonvolatile memory 3 can be used uniformly, and rewriting concentrates on a part of the area, thereby preventing reaching the rewritable upper limit of the nonvolatile memory. Can do.
If the power is cut off during entry update, the entry being updated is always at the end (see step A1), and the non-volatile memory usage area length stored in the external storage is the previous entry being updated. Points to the end. Therefore, at the next startup, in step E6, the entry part being updated is initialized, so that the consistency of the nonvolatile memory can be maintained.
Similarly, when adding data, if the power is turned off before updating the non-volatile memory usage size and checksum value, the variable being added (more specifically, the entry) can be invalidated. .

  When the data is deleted, if the power is turned off before updating the non-volatile memory usage size and the checksum value, a checksum mismatch occurs. This is because the data of Status 311 contains data “ADDED” in the character string. As a result, the checksum value is changed by the difference value due to the update to the data “DELETED” of the character string. Therefore, in such a case, in order to avoid simply judging that “a failure has occurred in the nonvolatile memory 3”, the checksum calculated by adding the difference value to the checksum calculated in step E4 is used. By performing this check again, it is possible to identify whether the checksum mismatch is due to a power failure during deletion or due to a failure in the nonvolatile memory 3. That is, if a match is obtained in the second checksum verification, it is determined that the power supply is being deleted. If a match is not obtained in the second checksum verification, it is determined that the non-volatile memory 3 is faulty.

  According to this embodiment, even if writing to the non-volatile memory 3 is interrupted due to a sudden power interruption or a FATAL failure, the data that has been interrupted does not cause a format error at the next startup. As described above, there is an effect that the consistency of the data in the nonvolatile memory 3 can be obtained.

  Even if the checksum value of the stored data in the nonvolatile memory 3 stored in the external storage device 2 cannot be updated due to a sudden power interruption or FATAL failure, as described above, The value difference that should originally be reflected in the checksum is limited to only the change in the contents of the status 312 of the entry, and the checksum is verified again by adding the difference. It is possible to distinguish the discrepancy due to the failure of the data and reliably diagnose the validity of the data in the nonvolatile memory 3.

  A program for executing at least a part of the processing of each component of the storage device according to the present invention by computer control and causing the computer to execute the above processing according to the procedure shown in each flowchart of FIGS. May be stored and distributed in a computer-readable recording medium such as a semiconductor memory, a CD-ROM, or a magnetic tape. A computer including at least a microcomputer, a personal computer, and a general-purpose computer may read and execute the program from the recording medium.

1 Nonvolatile memory management service (control mechanism)
2 External storage device (second storage medium)
3 Nonvolatile memory (first storage medium)
21 Non-volatile memory usage size 22 Checksum value
31 Entry.
311 Signature (valid / invalid / flag) storage item 312 Status storage item 313 Data length storage item 314 Data section (data storage item)

Claims (8)

A first storage medium for storing given data, attribute information of the data, and attribute information of the entry in an entry constituting one storage unit;
A second storage medium that stores a usage amount of the first storage medium to date and a checksum value for storage data in the first storage medium indicated by the usage amount;
A storage device comprising: control means for controlling storage contents of the first and second storage media based on a given command.   2. The storage device according to claim 1, wherein the data is variable-length data, and the control means includes information indicating a length of the data in the attribute information.   The control means indicates two types of information as attribute information of the entry: Signature information indicating whether the entry is valid or invalid, and Status information indicating the status of the entry. Alternatively, the storage device according to claim 2.   The control means indicates, as the Status information, a character string “ADDED” when an entry is newly generated, and a character string “DELETED” when the entry is deleted. The storage device described.   The control means calculates a checksum value of stored data of the first storage medium up to the present within a range of the amount of use up to the present stored in the second storage medium at the time of restart from interruption. And comparing the calculated checksum value with the checksum value stored in the second storage medium to verify the stored data up to the present time in the first storage medium. The storage device according to claim 1.   The control means calculates a checksum value of stored data of the first storage medium up to the present within a range of the amount of use up to the present stored in the second storage medium at the time of restart from interruption. The calculated checksum value is compared with the checksum value stored in the second storage medium, and the checksum value stored in the second storage medium if the comparison does not match The difference between the checksum values due to the difference in data between the character strings “ADDED” and “DELETED” is corrected, and then recompared with the calculated checksum value. The storage device according to claim 4, wherein when there is a mismatch, it is determined that a failure has occurred in the first storage medium.   The storage device according to claim 1, wherein the first storage medium is a nonvolatile memory. Storing given data, attribute information of the data, and attribute information of the entry in an entry constituting one storage unit of the first storage medium;
Storing, in a second storage medium, a usage amount of the first storage medium up to the present and a checksum value for storage data in the first storage medium indicated by the usage amount;
And a control step of controlling the storage contents of the first and second storage media based on a given command.

Images